Forecasting preanesthesia clinic appointment duration from the electronic medical record medication list

When scheduling clinic appointments, scheduling patients expected to have different visit durations for different minutes of time reduces patient waiting time and staff idle time. Maintaining an active medication list is an important (and, in the United States, required) component to the meaningful use of electronic medical records. We hypothesized that the count of medications from the medication list would be a better predictor of the time taken by a nurse practitioner to evaluate the patient preoperatively than the American Society of Anesthesiologists' (ASA) physical status and other demographic variables.

Using 69,654 preoperative visits, we obtained the number of different medications taken by the patient and demographic variables, including ASA physical status, ASA base units, and body mass index. For each independent variable, we applied transformations and calculated the Pearson correlation giving the largest correlation with the log(10) (duration), which followed a normal distribution.

Only 18% of the patients had been evaluated previously at the preoperative facility, making use of the prior ASA physical status ineffective for forecasting. The number of medications was a more accurate predictor of appointment duration than any of the other 8 variables (each Bonferroni corrected P < 0.0001), including ASA physical status.

Schedulers can use the number of medications that each patient is taking when choosing the time for preoperative evaluation. This approach can take schedulers only approximately 10 seconds extra per patient when scheduling the appointment.